Image Recognition System Based On The Gasochromic Sensor

As the surprising development of the computer technology, the information can be processed very well, but relatively the technology of gathering information is not good enough, it restricts the development of sensor. So here A new gas sensor was designed with sensitive arrays and embedded system. The sensitive arrays were made of several sensitive material metalloporphyrins and a new class of platinum vapochromic salts, which can change their colors contacting volatile organic compounds. And the embedded system was used to gather and process the array images. Then the Hough transform and color recognition method were used to distinguish the images. The system work on a new principle named Smell-seeing, and it's a new method for sensors to detect the gas. Now it can distinguish 14 volatile organic compounds perfectly such as some kinds of alcohols, amines, aldehydes, benzenes and ethyl acetate, THF.A new method called"smell-seeing"was found, that can distinguish volatile organic compounds. It made use of material metalloporphyrins as sensitive arrays that can change their colors contacting volatile organic compounds to detect gas. The method is very similar to using litmus paper to determine if a solution is acid by seeing if the paper goes from blue to pink, but to be much better, it includes more sensitive materials. Based on this principle, a sensor based on color changeable materials to organic gas was designed. The main research contents include:1. Have a research on a new method to detect volatile organic compounds, it includes the principle of color array's diversification and also a principle of image distinguishing. Then the steps of the experiment were designed.2. Make a gas sensitive array and complete the database with the vectors that can represent the gas. Here some materials sensitive to volatile organic compounds and how to make array with them were introduced, and then the arrays were tested to contact the volatile organic compounds so that the database was completed. Later, with the help of MATLAB, the distances of the vectors that represent gas information can be calculated and the advantages of the new method to detect gas were analyzed in the way of principle component analysis. According to the distances, the gas that our array can detect can be determined.3. Accomplish the programs to gather and save the images according to the need of the system.4. Design the method of image distinguishing. The main task is to design an efficient and simple method to get gas information according to the color array image. Here on the basis of the images, the edge detecting technology and Hough Transform were used to get the useful information of the image, and then with a pattern recognition based on statistics, the gas can be determined.5. Design the whole system. According to the need of detecting gas, the first part of the system is build up with USB camera and sensitive array fixed on the camera, and the second part of the system is OK2440 embedded system, which can gather and process the array images. After writing the corresponding processing programs with C language and building it to the embedded system, the system is accoplished.